Evolutionary approach for multi-objective optimization of wireless mesh networks

For a multi-hop wireless mesh network it is well known that network throughput as well as network energy consumption acts as two major performance metrics. Hence there is a real need to have a thorough study on how one can optimize both objectives simultaneously. In this paper, we take a evolutionary optimization approach to optimize the performance objectives. In order to stress on throughput and energy performance, we assume a simplified link layer scheduling by using orthogonal channels among the data links. We use the multiple reference point approach for the above multi-objective optimization problems which is intrinsically of discrete and combinatorial nature. During the course of approximating the Pareto Frontier, multiple reference points can be used instead of traditional methods. The multiple reference point approach has the benefit of parallelization by finitely partitioning the decision space.